In a germ free mouse there is minimal lymphoid infiltration in the gut. Upon colonization with normal flora there is significant influx and activation of lymphocytes into the gut mucosa and the phenomenon of controlled or physiologic inflammation is established. It is generally recognized that regulatory cells play a key role in the maintenance of this state. A trialogue between bacterial epithelial: T cell interactions has been demonstrated in a variety of different studies. We propose that bacterial:epithelial interactions promote an environment where specific Tregs can be generated. Commensal flora may upregulate/induce the expression of nonclassical class I molecules invovled in the activation of Tregs as well as promote selective chemokine production that recruits Tregs or Treg precursors into the mucosa and epithelium. Within the epithelium bacterial Ags presented by nonclassical class I molecules on lECs can activate Tregs to effect their function: the maintenance of the normal physiologic inflammatory state. In inflammatory bowel disease any number of defects in this process can lead to a failure in regulation. We have already shown that IBD lECs display defective expression of gplSO and CDld. The mechanism(s) underlying these defects have not been defined. Alteration in chemokine or cytokine production may account for these observations. This proposal will test distinct aspects of this model (in health and disease - all in human tissues). Specific Aim #1. Determine which Tregs are present (or absent) in the normal and IBD intestine and correlate these with defects in non-classical class I molecule expression in IBD (aim #2). Is the defect in TrE cells in IBD tissues reflective of a selective or global defect in Tregs? Specific Aim #2: Determine the expression and regulation of non-classical class I molecules on intestinal epithelial cells and define defects in expression on IBD lECs.